Micrsoscope photography: WARNING large pictures

[SiBurning]

Apologies for making the page so wide. I didn't want to resample the pictures and maybe ruin them.

There doesn't seem to be a thread about microscope photography going back a good while, so I started a new one. Thought it'd be fun to take close ups of razors, and see who else is doing it, and how.

I'm using a meiji stereo zoom with a camera trinocular port. It's a low powered "dissection" microscope. It'll go up to about 100X or so, although my lenses can only get to 45X. I picked up a cheap video camera to play around with. Don't know much about optics, photography, or microscopes.

It's kind of difficult to set up good lighting. Don't know which sources and holders are best. Also, a straight razor needs to be placed very precisely to have the entire visible length of the blade in focus. Just setting it down on the platform leaves it sitting at a slight angle, with one end of the blade higher than the other. That's enough to keep most of the visible part out of focus in these pictures. Of course, lighting plays a great part here. There's also no way to control things like fstop and exposure time on this video camera.

Was considering getting the adapter for my camera, a Nikon D40, but the adapter cost almost as much as the camera kit. I think the zoom feature requires special optics. Don't know much about how it all works, except that you need the right lenses to focus at the right distance. For now, I can take pictures with the cheap video camera by adjusting the head distance separately for the camera. At this distance, the eyepieces are out of focus. The video camera has 1.3 Mpixels. The camera has 6Mp.

Here's some pictures of a straight razor. Those long scratches might be from poor stropping using the paste side. One thing's apparent. There's a good bit of my face still on the razor, as well as leather particles from the strop.

Here's my trusty '54 Gilette relaxing after a day's work.

I originally picked up the microscope for electronics work.

We tried to solder a jumper across the switch that was pulled out. This picture is direct from a normal camera.

We managed to get a ball of solder on there, but it didn't make the electrical connection we hoped for.

Without the microscope, we couldn't see the wires connecting (or not connecting any longer) to the voice coil on this headphone.

 

[Sullybob]

Great pictures, thanks for sharing them. I have a cheap microscope and I find it very helpful to be able to clearly see the edge.

 

[Haiku]

I have to confess, while I'm not generally a fan of microscopic photograpny, I found all of your photos absolutely intriguing - thanks very much for posting them!

 

[arghblech]

Neat!

 

[jzono1]

Btw, for cheap edge inspection dealextreme got some neat 60-100x pocket microscopes that are good enough to inspect the bevel and about 20$ shipped.

 

[SiBurning]

Picked up a mount for a real camera which takes much better pictures. Thought I'd share some more pictures and the little I've learned.

To really see the actual edge takes something like an electron microscope with magnification in the thousands. I can't get over 90X with this thing--not with the camera anyway.

The main challenge is getting the pictures in focus. The camera's mounted to the eyepiece instead of the trinocular port because I'm missing a few parts from this scope. Mounted on the eyepiece, the camera shudder shakes the head enough to blur the pictures. I can see the image still shaking through the viewer after the shudder closes. I only hope it works better in the trinocular port once I can replace the missing parts. Otherwise, It'll take a more stable stand. A single piece boom stand isnt the most stable thing.

The choice of camera is also important. It's very difficult to focus through the viewer because it doesn't have enough resolution to see if the picture's in focus. There's some bad moire interference in the viewer, which makes it really hard to see the fine detail. This would be a lot easier with a camera lets you preview live on the computer. Unfortunately, my Nikon D40 won't do that.

So here's my little lighting experiment. These pictures are all of the same area, but with the lights (two fluorescents, one white balanced, the other a cheap desk lamp) held at different angles and distances. That edge is very hard to capture and I want to see what different lighting techniques can do.


large image


large image


large image


large image


large image


large image


large image

 

[Stubblefield]

Impressive. I 'd love to see a before and after shot on a blade, to see what one shave does to a blade.

Thanks for posting these!

 

[SiBurning]

Impressive. I 'd love to see a before and after shot on a blade, to see what one shave does to a blade.

Thanks for posting these!

Slowly working towards it, but it's a long way off. I can see the difference in the scope with bits of the edge rolled over after shaving, and then corrected after stropping. The stereo view helps a lot. I doubt I'll be able to reproduce that in photographs any time soon, partly because of the expense, partly because there's a lot to work out before being able to photograph at that level, and only then is it worth bothering with layers or a stereo picture.

Photographs need a lot more light, and that extra light causes shadows and reflections. That's the next challenge to tackle. You can see what I mean from the pictures. By the way, there are the original shots, in original resolution, converted to jpeg without any image correction.

Next time, I'll mark the spot and rephotograph after stropping and we'll see what comes out. This is my just my first day using the DSLR.

 

[fatpanda]

Thank you for doing this!!! I really want to see this.

 

[Stubblefield]

Photographs need a lot more light, and that extra light causes shadows and reflections. That's the next challenge to tackle. You can see what I mean from the pictures. By the way, there are the original shots, in original resolution, converted to jpeg without any image correction.

Lighting in the second and fourth pix looks good, and render with a bit of detail, tho I only see one grind on the blade. Are all thes shots from a straight?

Nice that you did not have to do any IC, those are nice images.

 

[SiBurning]

Yes, they're all straights. Each set is a different straight. The first post is a 5/8 Radio singing extra hollow. The previous post... I forget which blade that is, but all 7 pictures are of the exact same spot on the blade. Here's another set of pictures of a third straight http://img195.imageshack.us/gal.php?g=s224.jpg at an even lower resolution than the first set.

The detail's actually overdone in those light blue shots. The scratches aren't as deep as they look, but the light blue shots also show the depth of the edge a little better. In the dark blue shots, the scratches seem lighter than they actually are, and the edge shows as just shadow and highlight. The truth is somewhere in between. I played around with color & white balance, but these 7 raw shots together show the thing better.

This might require polarized light or filters. I'm reading up on these things, and might try a few, but there's not a lot of information on photomicrography (learned a new word this weekend) and most of it is for transluscent stuff like biological cells.

As for seeing the edge after a shave, there's not much to see. It's really just reflections of light off the edge. A while back, I noticed on one razor that they all went one way, so I stropped the razor heavily in that direction for a while, and that made the reflections go away, and gave me back a nice sharp edge. I do hope to be able to eventually photograph the reflections, but if my math is correct, actually seeing the edge damage in even the grossest detail is impossible with a light microscope. Those microfractures aren't much bigger than a wavelength or two of red light. It's fascinating that a leather strop can actually affect steel at that level.

 

[SiBurning]

First attempt to photograph using reflected light differential interference contrast (DIC), which shows a surface map of the specimen. On the downside, the depth of field is rather small with these lenses. These were taken using an Olympus BHM with 160mm reflecting DIC set & MPlan-N optics, and an Amscope 1.3Mp USB camera inserted as an eyepiece.

100x focused right on the edge 1280x1024

100x focused just off the edge 1280x1024
The focus is backed off the edge slightly and onto the surface material of the blade.

100x focused off the edge with different "shadowing" to capture the topography
1280x1024

100x focused on the edge with different "shadowing" to capture the topography (again)
1280x1024

100x focused on a larger area near the edge (Looks like I got it tuned better by this point)
1280x1024

10x focused at the bevel
1280x1024

10x at the final roll of the edge
1280x1024

100x again right at the edge
Just for fun, you can clearly see the lamp filament in the last 2 pictures--the yellow vertical lines on the right.
1280x1024

100x again just off the edge
1280x1024]1280x1024

 

[DunEdinRanger]

Great photos! Thanks for posting these.

 

[SiBurning]

Here's the microscope that took those last pictures.

It's an Olympus BH style (1970's type with 160mm optics) adapted with parts from a BHM and BHC to provide light from the top and bottom. The mounted nosepiece has the DIC set. Those objectives each sit on a contraption with a prism and a screw for adjusting the prism offset. The second nosepiece has Neo objectives that do brightfield & darkfield.

After several months of tinkering and replacing parts, it's finally somewhat functional, but still needs some work.

 

[lindcruiser]

Cool pics!

 

[SiBurning]

Someone pointed something out... The numbers in post 12 refer to the objective used, not the apparent factor of magnification. It means the objective was 100x, which when used with a typical 10x eyepiece would mean 1000x. The camera doesn't use an eyepiece, but the pictures are blown up, so at a guess it's somewhere over 100x.

I'm waiting on a measuring device, so the next set of pictures should have the actual dimensions and/or magnification factor listed.

 

[SiBurning]

Did some measurements and calculations...

The 100X pictures as posted here (cropped) measure 67 x 83 microns (width x height). That's 0.00264" x 0.00327". That's a magnification of 3000. Put another way, that's 176x218 lengths of violet light and 89x110 lengths of red. So don't read too much into fine details. It's probably mostly coatings anyway.

The 10X pictures are about 10X the size (687 x 858 microns). The uncropped 100X pictures measure 107 x 85 microns (1100 x 880 for the 10X).

 

[SiBurning]

100 microns along a used feather DE (about 1-2 hairs thick)

So what I'm attempting to show here is the best compromise I can make in a single picture to show all the feature of the blade.

1. To the right you can see the grinding marks on the bare metal, or at least on the harder, thinner innermost coating, if there is one.

2. To the left, there's another flat patch that seem to be a thicker layer of coating than what appears on the right. This is what makes me think there are multiple layers of coatings on the blade.

3. If you follow the left flat patch up to the very top of the picture, you can sort of make out that the grind marks extend into the thicker blob layer of coating.

4. In the center towards the front, you can see the topology of the blob layer of coating, as well as how it cracks when you shave with the blade.

5. Looking at the very edge of the blade by the flat patch on the right, you can see microfractures that are too small to be shown in detail with a light microscope.

6. The left half of the blade edge is covered with fragments from the shave--that's soap, skin, and maybe some hair. There's also the occasional metal fragment sort of extruding off the edge formed by parts of the coating.

I apologize for parts of this being so blurry, but that's the compromise that needs to be made in a single picture. I especially wanted to capture the 3D topology of the coating, while retaining the other features, and keep it to a single picture. Once I can get good pictures of a single plane, I'll try stacking to get the depth in focus. That's when this journey finally pays off.

For comparison, here's a (shrunken picture of a) 0.01 mm / line rule using the same camera and objective. Although shrunk, this picture should display the same areas as the picture above. With about 10 lines showing, the camera's taking a picture 100 microns across. This is roughly the same length of blade as the previous set of pictures, which were 67x83 microns, but the blades ran diagonally in those.

Technical notes...

Got a new head and the light's aligned in the center now. Also, it's trinocular, so the camera's mounted on the head instead of in the eyepieces. This will allow me to use my DLSR. For now, it's easier to use the software on the USB camera.

Theoretically, with the light aligned, there should be more detail in these pictures than in the previous set. It's may actually be harder to make things out this way. The off center light seems to have been showing some kind of artificial details by way of having too much contrast.

 

[SiBurning]

Tried stacking to get more of the image in focus.

Stack of 21 images using DIC, applying 20% contrast threshold. This is showing appx 580 microns of blade length.

Removed the polarizers and stacked 18 images, applying 50% contrast threshold. That thing hanging off the bottom right is a chunk of coating stuck in that blob of... whtever it is.

Here's another shot of the same area for some perspective. The primary bevel is 0.5mm deep, the secondary maybe 1/5 of that. Note how the pictures above are entirely within the secondary bevel.

 

[ambrose]

very cool